It has been known that corrosion of amorphous carbon catalyst supports and metal catalyst, which occurs during startup and shutdown of polymer electrolyte membrane (PEM) fuel cells, results in a permanent decay of fuel cell performance. It has also been known that the corrosion is due to a reverse current situation in which the cathode potential may be well in excess of one volt higher than the potential of a standard hydrogen electrode. It is believed that this is caused by both hydrogen and air being present at different locations within the anode flow field. During a shutdown period, unless an inert gas purge is used, air will slowly, uniformly fill both the anode and cathode flow fields of the fuel cell. During startup, hydrogen is fed to the anode flow field which results in the inlet to the anode flow field being primarily hydrogen while the exit of the anode flow field is primarily air. An electrochemical reaction occurs between the fuel rich zone in the anode flow field and the oxygen rich zone in the anode flow field that causes the potential of the anode in the oxygen rich zone to increase to the air open-circuit potential. This in turn raises the potential of the cathode, opposite to the air rich zone on the anode, to a potential of 1.4–1.8 volts versus a standard hydrogen electrode. This potential causes the carbon based catalyst support to corrode and results in decreased cell performance.
In copending U.S. patent application Ser. No. 10/305,301 filed Nov. 26, 2002, U.S. Publication No. US2003-0134165 A1, it is shown that as the fresh hydrogen-containing fuel flows through the anode flow field upon startup, to displace the air therein, the corrosion of the platinum catalyst and catalyst support occurs as the hydrogen/air interface moves through the anode flow field. The extent of corrosion is mitigated by rapidly purging the air with hydrogen during startup of the fuel cell. In a similar fashion, it is known that as purge air is passed through the anode upon shut-down, there is a hydrogen/oxygen interaction, which creates a potential safety hazard and may cause undesirably large voltage excursions in the cells, as described in copending U.S. patent application Ser. No. 10/305,300 filed Nov. 26, 2002, U.S. Publication No. US2003-0134164 A1.
In automotive applications, that may experience 50,000–100,000 startup/shutdown cycles, this results in catastrophic performance loss. Heretofore, solutions to this problem include stabilizing the fuel cell stack by purging the anode flow fields with an inert gas, such as nitrogen, and maintaining an auxiliary load across the fuel cell stack during the shutdown and startup processes.
In automotive applications, the availability of an inert gas, and the apparatus to employ it for purging will be prohibitively complex and expensive. The use of an auxiliary load requires dissipation of the heat generated thereby, which may typically occur in a reservoir of a water circulation system or a coolant system, or may occur with air cooling.
Referring now to FIG. 1, a vehicle 150 includes a fuel cell stack 151 comprising a plurality of contiguous fuel cells, each having a membrane electrode assembly (MEA) 16, only one fuel cell 12 being shown in FIG. 1. The electrical output at the positive and negative terminals of the fuel cell stack 151 is connected by a pair of lines 155, 156 through a switch 158 to a vehicle propulsion system 159. The output is also connected through a switch 160 to an auxiliary load 161 in a reservoir 164 of a water circulation system, the reservoir having a vent 165. The water circulation system may include a trim valve 166, water passages, such as those within water transport plates 84, 86, 89, a radiator and fan 168, 169 which is selectively operable to cool water circulating in the system, and a water pump 170. Ambient air at an inlet 173 is provided by a pump, such as a blower 174, to the oxidant reactant gas flow fields of the cathode 19, and thence through a pressure regulating valve 175 to exhaust 176. Hydrogen is supplied from a source 179 through a flow regulating valve 180 to the fuel reactant gas flow fields of the anode 17, and thence through a pressure regulating valve 181 to exhaust 182. A fuel recycle loop includes a pump 183.
A controller 185 responds to load current determined by a current detector 186 as well as to the voltage across the lines 155, 156; it may also have temperature of the stack provided on a line 187. The controller, in turn, can control the valve 180 over a line 190 as well as controlling the other valves, the switches 158, 160 and the pumps 174, 170, as shown in FIG. 1.
The controller 185 responds to start and speed control signals from the vehicle propulsion system 159 on lines 193 and 194, which will indicate when the fuel cell should commence operation, and the amount of power being demanded by the vehicle propulsion system. Whenever a startup signal is sent from the vehicle propulsion system 159 over the line 193 to the controller 185, signals from the controller will cause the valves 180, 181 and the pump 183 to be operated appropriately so as to provide fuel reactant gas to the flow fields of the anode 17, and the valve 175 and pump 174 will be operated appropriately to provide ambient air to the flow fields of the cathode 19.
When fuel and air of sufficient quantity have been provided uniformly to the cells, open circuit voltage will be detected on the lines 155, 156 by the controller 185. At that time, the controller may close switch 160 so as to connect the fuel cell stack 151 to the auxiliary load 161 in the reservoir 164 and may also close the switch 158 so as to connect the fuel cell stack 151 to the vehicle propulsion system 159 at the same time, or later.
Whenever a shutdown signal is received from the vehicle propulsion system 159, the switch 160 will be closed so as to connect the auxiliary load 161, as the switch 158 is open so as to disconnect the vehicle from the fuel cell power plant.